From the course: AI Workshop: Hands-on with GANs with Deep Convolutional Networks
Unlock this course with a free trial
Join today to access over 23,300 courses taught by industry experts.
Deep convolutional GANs (DCGANs)
From the course: AI Workshop: Hands-on with GANs with Deep Convolutional Networks
Deep convolutional GANs (DCGANs)
In this movie, let's understand what a deep convolutional GAN is and how it's different from a regular GAN. A deep convolutional generative adversarial network is a GAN where each adversary is a convolutional neural network, and these CNNs have certain architectural constraints. The architectural constraints are so set up that they help the adversarial pair learn a hierarchy of representations which help create better-generated images. Now, deep convolutional GANs build upon the original GAN paper. They use the same principle as GANs, the generator and discriminator as adversaries. The only difference is in the architecture for the neural networks that we use for the generator and the discriminator. Both neural networks are architecturally constrained convolutional neural networks. The constraints introduced in these architectures help the CNN be more stable during the training process of the GAN. The authors of the paper on deep convolutional GANs found that it was hard to use…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Contents
-
-
-
-
-
Generator and discriminator1m 43s
-
(Locked)
Deep convolutional GANs (DCGANs)6m 16s
-
(Locked)
Setting up data for GAN training4m 8s
-
(Locked)
Setting up the generator and discriminator6m 18s
-
(Locked)
Output from an untrained generator and discriminator2m 17s
-
(Locked)
Setting up the GAN training loop7m 9s
-
(Locked)
Viewing GAN training results4m 34s
-
-